[NO PLAYERS] Seeding and ranking discussion

[The Black Sword]

This world will have a serious supply of excess glue.

[GermanJoey]

In my experience it's a pretty unpopular wonder that rarely falls before 9-10 cities in veteran games.

Well, I'm exaggerating a bit of course (I almost always exaggerate, which perhaps is itself an exaggeration! ) but, there have been a lot of... questionable HG plays the last couple of years. I've been curious about this, so I looked at the last 10 or so pitbosses and PBEMs, not counting PB24 or SMEGmod PB26, and here's the stats:

(PB27 and PB29 still ongoing, but the HG landed many months ago in both of these games so I don't think I'm spoiling anything)

PB21: Barteq, 9 Cities (unreported, but from memory) (ToW, Stone; 300h, 9*30 - 300 = -30h net loss)
PB22: HAK, 5 Cities (RtR, Exp, Stone; 250h; 5*30 - 250 = -100h net loss)
PB23: Whosit, 11 Cities (RtR, Exp, Stone - 250h; 11*30 - 250 = +80h net gain)
PB24: N/A
PB25: Gavagai, 9 Cities (RtR, Exp, Mono, Stone - 229h; 30*9 - 229 = +41h net gain) (these all went into Numids so no +25% bonus on the output ) (and IIRC, HAK wanted to build the HG at 7 cities in this game but Mindy argued him out of it)
PB26: MOD TOO CRAZY TO COMPARE, although IIRC Pindicator landed a good one for 13-14 pop.
PB27: Cheater Hater, 7 Cities (RtR, Exp, Stone - 250h; 30*7 - 250 = -40h net loss)
PB28: Dreylin, 11 Cities (RtR, Exp, Stone - 250h; 30*11 - 250 = +80h net gain)
PB29: TBS, 13 Cities (ToW, Ind, Stone - 213h; 37.5*13 - 250 = +237h net gain)
PB30: Dantski, 6 Cities (RtR, Exp, Stone - 250h; 30*5 - 250 = -70h net loss)
PB31: N/A, just started
PBEM66: Couldn't figure it out
PBEM67: very very late build by DTG, like T140 standard speed, probably for like a billion pop
PBEM68: Retep built it, so unreported and who knows when it was built
PBEM69: HAK, 8 Cities (BTS, Stone - 250h; 30*8 - 250 = -10h net loss)
PBEM70: too recent, but it was pretty big
PBEM71: N/A, unbuilt so far AFAIK
PBEM72: N/A, just started

So, out of 11 games, we have 5 where it was a net loss, 3 where it was a profitable build, and 3 where it was a very very profitable build.

On your other comments, I completely agree with you. It's one of the funnest wonders to land, maybe the single most fun, but once you actually look at the opportunity cost and risk, it's very very expensive. It's also not like the Mids or something, where it can allow you to completely change the way you play, because it gives a very specific, one-time benefit.

First of all, like you say, not all pop points are created equal. +1 pop in a happy-capped city is pretty worthless, as is +1 pop in any city that you're not willing/able to whip if it doesn't have a new improved tile to work. On top of that, if this were an RtR game, you may not even be getting 30h per whip! To be fair, +1 pop can also be worth a lot more than just 30h - for example, a BTS Exp civ can whip a granary in a brand-new city, or sometimes that +1 pop lets you whip a monument right away in a city that wouldn't really work without one, or it could be +1 grassland cottage in a Fin capital, etc.

Next, like you also say, the aqueduct greatly increases the risk of this wonder, because if you fail it then that's 100 less failgold. In fact, you have to fail the wonder at almost the point of completion for the failgold to even be better than building wealth for the duration, and that's *with* stone. (i.e. 250h total for 300g) Dantski has Hammams here, which are a hell of a lot better than a regular aqueduct for sure and thus dampen his risk a bit, but they are still fairly situational builds. He definitely wants one in his capital, eventually, but it seems like he has bigger problems than not growing that city onto cottages fast enough considering that his capital is red-faced at size 9 in his screenshot. Then there's the temptation to whip the Aqueduct. This can be disastrous for your net profit. In BTS, one could very easily triplepop whip the thing, potentially even quadruple. That's a hell of a lot of hammers off the net cost, but then you need to subtract this whipped pop from the free pop. For example, a 13 pop HG that used a quadruplepop whip on the aqueduct (or a doublepop on an aqueduct and a doublepop on a settler the turn before, to form an overflow chain) really only gained 9 pop.

Finally, you've got your worker costs. I never thought about workers being part of a wonder cost until relatively recently, but they totally are. For example, lets say you want to put 6 chops into the HG. However, 6 chops is 24 worker turns at minimum, perhaps even more if you road some of the forests or do prechops. 30 worker turns for 6 chops wouldn't be unreasonable for base-cost 400h wonder. So what? Well, 30 worker turns in the ancient era is actually a huge amount, way too much to just sneak in the margins. You're either going to fall behind on improving your tiles or else have to build 2-3 additional workers than you would otherwise need *just* to chop out this damn thing in a reasonable timeframe (again, important because the risk of this wonder is so high). Extremely mild ancient-era PB29 spoiler:

In PB29, I ran some late ancient-era sims to see how an HG timed in the early T90s would work out vs not trying for it at all, and I was surprised to learn that my total CY at a date 10T into the future was the same whether I built the HG or not! I eventually figured out that it was in a large part because of this extra worker cost - I needed a lot more workers early for the same number of total improvements, and furthermore some of the regularly-scheduled ones needed to be reallocated, and thus my land wasn't as improved as efficiently. Also, whipping the Aqueduct hurt too. So I didn't really try for it, and at any rate TBS ended up finishing it before I even had Math, IIRC.

Finally, there's the fact that you have 250-300 hammers sitting in the queue forever, instead of other things. In the simplest case, this could be 3 settlers or so, settlers that could plant many turns before the HG completes, gain land, start growing, etc. If you're out of land, then the opportunity cost of the HG is a lot lower, but giving up ~40-50 tiles probably is not.

[The Black Sword]

The GE points are probably worth about a pop point too.

[GermanJoey]

True, true.

[Dreylin]

OK, so I didn't see 'wolf's post before he edited it out, but from Gawdzak's reaction he's building GLH.

CivStats doesn't have him with a score bump in the last couple of turns and no-one has reported seeing it in the notification log, so I assume that means he just needs to log back in and change some builds before the turn rolls ... in which case no replaying needed, right?

[Commodore]

He chopped it down to being due in 2t. Kind of sucks.

[Dreylin]

Ah, so he doesn't even need to log in since he can fix it next turn - although it would be better if he could so the chop hammers wouldn't be wasted.

Ipecac has played since, and there's undoubtedly some combat involved there, so it's probably best to avoid a reload. 'Wolf eats the invested hammers for not paying attention.

Fake Edit: I see he's logged in.

[spacetyrantxenu]

I believe this is what you call chickens coming home to roost.




That's a lot of chickens.

[Elum]

They that sow the wind.

Molach's galley has a nice view!

[Commodore]

Er, so is TheWannabe going to hit space before the rest hit Industrial?